Relationship of Dengue with Malaria and other social and climate covariates from four different cities of Pakistan

The rational of this study is to highlight the possible relationship of Dengue with Malaria and other social and climate covariates from four different cities (Lahore, Karachi, Islamabad and Rawalpindi) of Pakistan. For this purpose the relationship between the occurrences of Dengue and Malaria, Dengue and flooding, Dengue and population, and Dengue and travelling in the study areas have been taken into account. Generalized Linear Mixed Model (GLM) with Markov Chain Monte Carlo (MCMC) algorithm is computed to see the random effects of different social (population, travelling, and malaria) and climate (minimum-maximum temperature, and rainfall) covariates on Dengue occurrence. Neural Network with Multilayer Perceptron is used to analyze the normalized importance of different covariates relative to Dengue occurrence. Results show that flooding, travelling, population and occurrence of Malaria are affecting the occurrence of Dengue in the study areas. Change in occurrence of Malaria is affecting the occurrence of Dengue as much as 5.4 times, whereas GLM with MCMC also shows significant random effects of Malaria, population and rainfall on the Dengue occurrence during the studied years (2009-2012). Key Words: Dengue occurrence, Malaria, social covariates, climate covariates, Linear Mixed Model and Neural Network